debugging-wizard
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a standard, hypothesis-driven debugging workflow. A comprehensive review of the instructions and examples found no malicious intent, obfuscation, or unauthorized capabilities.
- [COMMAND_EXECUTION]: The skill provides reference commands for legitimate, local development tools such as Python's pdb, Node.js's inspector, Git's bisect, and Go's delve. These commands are informational and used for local execution within the developer's environment.
- [PROMPT_INJECTION]: The instructions do not contain attempts to override safety guidelines or jailbreak the agent. The constraints provided are strictly related to debugging best practices (e.g., 'reproduce the issue first', 'do not guess without testing').
- [DATA_EXFILTRATION]: No network exfiltration patterns or hardcoded credentials were identified. The external links provided point to legitimate documentation hosted on GitHub Pages.
Audit Metadata